Artificial Selection on the GA2ox Gene Family Contributes to Plant Architecture Improvement in Upland Cotton

Gibberellins (GAs) play a crucial regulatory role in the growth and development of cotton (<i>Gossypium hirsutum</i> L.). Through bioinformatics analyses, we identified a total of 39 <i>GA2ox</i> genes (encoding gibberellin 2-oxidases) in the cotton genome, designated <i>GhGA2ox1</i> to <i>GhGA2ox39</i>. Based on phylogenetic analysis, these genes were classified into five groups. We further examined their gene structures, conserved motifs, and chromosomal distributions, revealing that members within the same group shared similar structural and motif organizations. Collinearity and cis-element analyses provided important insights into the evolutionary history and regulatory potential of the <i>GA2ox</i> gene family in cotton. Notably, using nucleotide diversity (π) and population differentiation (<i>F_ST</i>) analyses across the entire family, we screened and identified nine candidate genes that underwent strong artificial selection during cotton domestication and improvement. Further haplotype-phenotype association analysis identified <i>GH_D09G0919</i> (<i>GhGA2ox31</i>) as a key regulator of Plant Height (PH). To validate their regulatory roles, we analyzed the genotype distribution in accessions with extreme phenotypes. The results revealed divergent selection histories for these two loci: the favorable allele of <i>GH_D01G0720</i> (<i>GhGA2ox23</i>) was already fixed in the tested population, whereas <i>GH_D09G0919</i> maintained significant natural variation. Specifically, the Hap2 allele of <i>GH_D09G0919</i> was significantly enriched in the shortest accessions compared to the tallest ones. Importantly, quantitative real-time polymerase chain reaction (qRT-PCR) analysis confirmed that the Hap2 allele drives significantly higher gene expression in leaves, suggesting that enhanced GA catabolism underlies the compact phenotype. Additionally, transcriptomic profiling revealed the tissue-specific expression patterns of candidate genes, implying their functional roles in development. Furthermore, functional validation using the <i>Arabidopsis</i> mutant of the homologous gene (<i>AtGA2ox8</i>) confirmed its conserved role in regulating plant height, as the mutant exhibited a distinct short-stature phenotype. These results uncover valuable genetic resources for molecular breeding to shape compact cotton architecture. Collectively, this study aims to analyze the evolutionary patterns of the cotton <i>GA2ox</i> gene family and to identify key genes that regulate plant height under artificial selection, providing theoretical support for molecular breeding of compact plant types.